Terms.
Framework agreements locally and internationally, with multi-year and volume terms, so every division buys on the same conditions.
Framework agreements are available locally and internationally, so divisions worldwide purchase under identical conditions instead of re-negotiating the same product four times. The model is a transparent price per seat: no metering, no token pricing, no maintenance fee, and updates included. Support comes with every commercial license. Analysis runs on your own infrastructure, independent of any coding agent, FPGA vendor, simulator or EDA suite.
Framework agreements · Price per seat, no metering or token pricing · Updates included, no maintenance fee · Support with every commercial license · Analyzed locally, on your own infrastructure
Framework agreements locally and internationally, with multi-year and volume terms, so every division buys on the same conditions.
A transparent price per seat. No metering, no token pricing, no maintenance fee, and updates included.
Node-locked file or internal FlexNet server, installed as a VS Code extension. Air-gapped operation is possible.
Included with every commercial license, with a dedicated account team and local assistance across eleven regions.
Not bundled into an EDA suite. No dependency that makes the next renewal harder to negotiate.
A buying decision for engineering tooling is rarely made once. It is made in one division, then re-argued in the next, under different terms, in a different currency, with a different procurement contact, until the same product is held on four contracts that expire in four quarters.
Framework agreements are available locally and internationally, so every division can purchase on the same conditions regardless of where it sits. One set of terms, one renewal conversation, one commercial relationship to manage, and no division negotiating against the position another division already agreed.
Multi-year agreements are available, and volume discounts apply. Combined with a framework agreement, that is what lets a large organization hold one price structure across divisions rather than discovering that two sites bought the same product at different rates in the same year.
Under a multi-year agreement the price per seat is fixed for the term. Seats added while the agreement runs are added at that same price for the rest of it, so a division that grows in month seven is not re-quoted at a rate nobody planned for. For a procurement function that is the difference between a forecast and an estimate.
That matters more than it sounds. The cost of inconsistent terms is not the price difference. It is the administrative load of maintaining several relationships for one product, and the loss of leverage that comes from fragmenting your own demand.
Sigasi Visual HDL is one extension with one semantic engine. The license key decides how much of it is unlocked, across three editions: Designer, Professional and Enterprise.
Each edition includes everything in the edition before it, so moving a team up is a key change rather than a migration or a reinstall. The edition in use is shown inside the product, and the product prompts when something needs a higher one, which keeps entitlement visible to the people using it rather than only to the license server.
Every Enterprise purchase also includes an Enterprise CI license, which covers automated use outside the IDE, in a build pipeline. That is worth knowing at quote stage, because automated and CI use is commonly the line item that gets priced separately elsewhere.
The full feature-by-edition breakdown is on the edition comparison page, and the three commercial routes are set out below. Confirm current entitlements and CI deployment terms for your team before purchase.
Evaluate it on your own project, use it free for academic and non-profit work, or license it for commercial development. The same semantic engine is behind all three.
Free, time limited
Install the extension in VS Code, point it at your own SystemVerilog, Verilog, VHDL or mixed-language project, and see what the semantic engine finds on real libraries, compile order and hierarchy.
All commercial use
The deterministic HDL platform for professional ASIC and FPGA teams. The same semantic engine serves engineers in VS Code and the coding agent your team already uses.
Designer
Semantic analysis, linting, navigation, and language-aware editing for professional RTL work.
Professional includes Designer
Extend development with graphical views and verification-oriented capabilities.
Enterprise includes Professional
Bring the shared semantic engine to the command-line and automated engineering workflows.
Free, academic and non-profit
The Enterprise feature set, free of charge, for students, instructors and other non-commercial work. There is no project size limit, so it holds up on a real design rather than only on a classroom example.
The price list is not public. The model is, and the model is the part that usually hides the surprises.
It is a price per seat. Within an edition there is no metering: nothing is charged per analysis, per file, per project or per run, and there is no token pricing that turns next quarter’s invoice into a forecast instead of a number. An engineer with a seat uses what that edition unlocks, as much as they need, which is what makes it realistic to give a seat to everyone who writes RTL rather than rationing a few.
Updates are included. There is no separate maintenance or support fee layered on top, and no point at which staying current becomes its own negotiation.
For scale: Sigasi is not priced like the advanced-node implementation and sign-off tools a CAD budget is used to absorbing. This is front-end tooling meant to sit on every RTL engineer’s desk, and it is priced so that it can.
This matters more as AI enters the flow. Usage-based and token-based pricing moves the cost of a tool from a line you can plan to a number that depends on how much your engineers and their agents happen to do that quarter. A per-seat model with no metering does not have that property, and a procurement function can forecast it.
Most of the lines in a cost-of-ownership model for tooling like this are the ones that are not in the license. Here they are in one place.
| Cost line | What you pay |
|---|---|
| License | Per seat, per edition, set in a quote for your team size |
| Maintenance fee | None |
| Version upgrades | Included |
| Support and user manual | Included with every commercial license, at every edition |
| Analyses, files, projects, runs | Not metered |
| Token or usage pricing | None |
| CI and automated use | Enterprise CI license included with every Enterprise purchase |
| License server | Node-locked file, or a FlexNet server inside your own network |
| Rollout | A VS Code extension, through the mechanism you already use |
| Seats added mid-term | At the agreed price, for the rest of the agreement |
The quote fills in the first row. Every other row is settled before the conversation starts, which is the point of publishing the model rather than the price.
Agent spend has become its own line in an engineering budget, and it is the one nobody can predict, because it is driven by how many attempts a task takes rather than by how many people are working.
Sigasi affects that line in two ways. Its own cost does not move with usage, because it is not metered. And an agent working against a project model that already answers where a type is declared, what a library unit resolves to and what a change touched has less reason to guess, retry and be sent back. Fewer blind iterations is the mechanism, and it is the same mechanism behind the engineering time saved, which is why both land in the same calculation.
Commercial use runs from a node-locked license file or from an internal FlexNet license server, whichever suits how your sites are organized. There is no requirement to reach a Sigasi-hosted licensing service during normal operation.
The analysis engine can operate without external internet access, provided installation, licensing, dependencies and any chosen coding agent are available inside the environment. For programs that work in disconnected or restricted networks, that is usually the question that decides whether a tool can be used at all.
It installs as a VS Code extension, which means deployment follows whatever mechanism your organization already uses for VS Code. There is no separate IDE to roll out, qualify and support.
Sigasi analyzes HDL locally, on the engineer’s workstation or inside your controlled environment. Project files are not sent to Sigasi servers for semantic analysis.
Telemetry is optional for commercial licenses and does not transmit HDL source code. It can include usage and technical metadata, so a security team should inspect the local log and the settings rather than take the sentence on trust. The free Community Edition does require it, which is one of the reasons it is limited to non-commercial use.
A connected coding agent is a separate component with its own provider, configuration and data handling, and must be reviewed on its own terms. Disabling Sigasi telemetry does not control Microsoft’s VS Code telemetry or any third-party extension. Keeping those three things separate in a security review is the difference between a clean assessment and a stalled one.
Professional support and the user manual come with every commercial license, independent of edition. There is no support level to negotiate separately, no edition that leaves you without help, and no maintenance fee to keep it active.
Accounts are served by a dedicated account team rather than a queue that has to be re-briefed every time something goes wrong.
Local assistance covers the USA and Canada, Germany and the DACH region, the UK, Ireland and the Nordics, India, Türkiye, Israel, Taiwan and South Korea, Japan, and Benelux, France and the rest of Europe. Those regions are handled by Sigasi in collaboration with local partners where available, so a division gets someone in its own time zone and language without the commercial relationship fragmenting into separate contracts.
Sigasi is not tied to one coding agent, one FPGA vendor, one simulator, or one vertically integrated EDA suite. It is model-neutral and toolchain-independent.
That has a direct commercial consequence. A tool bundled into a larger suite makes the next renewal of that suite harder to negotiate, because unpicking one component means unpicking the relationship. An independent layer does not create that dependency, and it keeps its value if your simulator, your FPGA vendor or your AI strategy changes.
The same applies to the pricing model. A per-seat license with no metering does not create a dependency that grows quietly with usage, which is the mechanism by which a cheap-looking tool becomes an expensive one that is hard to leave.
It also survives disagreement between programs. One may approve a cloud coding agent, another an internally hosted model, another none at all. The same deterministic layer serves all three, so the tooling decision does not have to wait for the AI decision.
Sigasi nv has worked on semantic analysis for SystemVerilog, Verilog and VHDL since 2008, close to two decades on the same problem. The technology is patent protected under EP 2175387 and US 8881074.
International customers outside the euro zone are invoiced in US dollars.
The software license agreement is public and can be read before any conversation, rather than arriving at the end of a procurement cycle. Sigasi operates a published responsible disclosure policy with a five business day acknowledgment commitment, documented on the security page.
A one-page procurement fact sheet carries the terms, licensing, deployment, data handling and support scope on a single sheet, for the people who need the summary in the file rather than the argument around it.
The comparison is not with another tool. It is with doing the work by hand.
Today an engineer establishes where a type is declared, which library unit a name resolves to, what a port change touched and whether an unfamiliar block does what it appears to do, by reading, grepping and asking a colleague. That time is already being spent and already being paid for. It does not appear on any budget line, which is the only reason it looks free.
The second half of the same cost is what happens when it is established wrong: a width mismatch or a stale dependency found in simulation, in CI or at integration bring-up rather than in the file where it was written, and the iteration that follows.
So the question a cost comparison has to answer is not whether the license is cheaper than an alternative product. It is whether it costs less than the hours it displaces.
Price is one side of the decision. The other is the engineering time that currently goes into avoidable work before a design reaches verification.
The calculator takes your team size, your cost per engineer and your project load and puts a figure on that, which is the number worth bringing to an internal approval rather than a vendor’s. Because the license is per seat with no metering, the cost side of that comparison is a figure you already know rather than an estimate that moves with usage.
Sigasi quotes a typical project moving from 38 to 28 weeks, based on feedback from large customers. Treat it as an illustration rather than a commitment: individual results vary by project, which is exactly why the calculation should be run on your own inputs before it appears in a business case.
Sigasi does not generate RTL and it does not decide. It does not replace simulation, formal verification, coverage analysis, synthesis, sign-off, or engineers. It does not certify a design or guarantee compliance, and no tool that catches issues early can promise to catch every issue. Capabilities vary by edition.
Stating this plainly is deliberate. A procurement file built on claims the product cannot carry is a problem that surfaces later, in front of the people who approved it.
A trial license is free and time limited, runs on the Enterprise Edition, and can be extended when an evaluation needs longer. The license file arrives by email within the next business day, and a field application engineer will set it up with you at no charge, on a real project rather than a demonstration design.
A team trial of two to three weeks on one active project is enough to produce numbers of your own: avoidable issues caught before simulation and CI, review churn in iterations and time to merge, onboarding speed, and surprises at integration bring-up.
For academic and non-profit work, Community Edition unlocks the Enterprise feature set free of charge, for non-commercial use only, requiring Talkback, without the standalone headless CLI, and with no project size limit.
Request a trial, book a free session, or open the pricing and contract conversation using the form below.
Tell us the team size and the editions you need. Framework terms covering several divisions are worth raising here rather than at the end.
Price is one side of the decision. The calculator takes your team size, your cost per engineer and your project load, and puts a number on the engineering time that goes into avoidable RTL iterations. Bring that number to the conversation.
Open the savings calculatorYour code stays with you
Sigasi analyzes HDL locally. Project files are not sent to Sigasi servers for semantic analysis.
Building HDL tooling since 2008
Sigasi has worked on semantic analysis for SystemVerilog, Verilog, and VHDL for close to two decades.